356, RIKARD STERNER 



rences which lie most inland and highest above sea-level fall about at the level 

 which corresponds to the situation of the shore about the time of the maximum 

 extension of the Ancylus Lake. 



If we summarize what has been said about the distribution-conditions of Cynan- 

 chum vincetoxicum in the south-east of Sweden, we get the following: — 



In Smaland and Ostergotland the occurrences situated furthest up-country 

 would be the oldest habitats of the species in those districts, and be survivals 

 from the distribution on the coast which existed at the time of the maximum 

 extension of the Ancylus Lake. Outwards from these oldest occurrences the 

 species spread itself abundantly as suitable new localities came into existence 

 with the gradual rising of the land, but, owing to the nature of the localities, in 

 conjunction with the pollination biology of the species, it did not spread inwards 

 over the mainland. In Sodermanland and Uppland, which for the most part 

 became land considerably later, the spread of the species inland seems to be 

 determined by the distribution of suitable escarpments and perhaps, so far as 

 Uppland is concerned, by certain human action on the vegetation and the 

 geographical development of the land. 



The fact that the xerothermic Cynanchum immigrated into the south-east of 

 Sweden about the time of the maximum extension of the Ancylus Lake — that 

 is to say, during what Sernander calls the Boreal period — thus stands in 

 agreement with previously expressed views concerning the history of the immigra- 

 tion of Swedish flora; see, for instance, Sernander 1894, p. 81 (cf. 1908, p. 219!); 

 G. Andersson 1896, p. 41; Sundelin 1917, p. 285; 1919; v. Post 1920; Sande- 

 gren 1920; etc. 



As has already been hinted, however, a number of species exhibit similar 

 distribution in the south-east of Sweden. These species are entirely dissimilar 

 to Cynanchum in many respects with regard to their biology and ecology. 



I have already given an account of the distribution of Melampyrum nemorosum. 

 Its present most inland occurrences might be regarded as remains of the original 

 distribution along the coast; but its inland limit is still unexplained. 



Sedum album, distributed in Central Europe and Scandinavia, shows great 

 resemblances to Cynanchum with regard to its distribution in the cast of Sv^eden. 

 But its Scandinavian distribution is in olher respects very unlike that of Cynan- 

 chum. It has an abundant and extensive distribution-area in the north- iaest of 

 South Sweden and in the south of Norway, an area which is connected with 

 its habitats in south-eastern Sweden by certain sporadic occurrences on the shores 

 of the lakes in Central Sweden. The biology and ecology of this species would 

 also seem in several respects to be unlike that of Cynanchum. The continental, 

 but in Middle Europe widely distributed Scutellaria hastifolia (Plate 10) — which 

 ought to be included amongst the flora of shore-meadows, which Sernander calls 



